Emotional tagging retroactively promotes memory integration through rapid neural reactivation and reorganization

Abstract

Neutral events preceding emotional experiences are thought to be better remembered by tagging them as significant to simulate future event predictions. Yet, the neurobiological mechanisms how emotion transforms initially mundane events into strong memories remain unclear. By two behavioral and one fMRI studies with adapted sensory preconditioning paradigm, we show rapid neural reactivation and reorganization underlying emotion-tagged retroactive memory enhancement. Behaviorally, emotional tagging enhanced initial memory for neutral associations across the three studies. Neurally, emotional tagging potentiated reactivation of overlapping neural traces in the hippocampus and stimulus-relevant neocortex. Moreover, it induced large-scale hippocampal-neocortical reorganization supporting such retroactive benefit, as characterized by enhanced hippocampal-neocortical coupling modulated by the amygdala during online processing, and a shift from stimulus-relevant neocortex to transmodal prefrontal-parietal areas during offline post-tagging rest. Together, emotional tagging retroactively promotes associations between past neutral events through stimulating rapid reactivation of overlapping representations and reorganizing related memories into an integrated network.

Threat learning impairs subsequent associative inference

“…Instead, we reasoned that the sample size should be (1) comparable to those reported in key studies that inspired this experiment, and (2) sufficient to reliably estimate parameters in a two-level model. Bisby et al consistently demonstrated impairments in emotional associative memory across three experiments with sample sizes between 17 and 27 21 , whereas enhancements of neutral memories following threat learning have been shown in samples of 30 participants 12 , 13 . We thus sought to include at least 30 participants, such that both an impairing or enhancing effect of threat on associative inference could be detected.…”

“…In line with this idea, specific retroactive emotional enhancements have previously been demonstrated: recognition memory was enhanced for neutral items that later, through new learning, acquired emotional significance as instances of a semantic category which predicts a mild shock 12 . As for associative memories, Zhu et al 13 recently revealed that when one element of an existing memory is paired to an emotional stimulus, associations within the original memory are strengthened. Similar evidence comes from a study using monetary reward as reinforcing stimulus, showing that a hippocampus-dependent mechanism allows positive value to spread to reactivated memories, thereby subconsciously biasing subsequent decision making 14 .…”

Threat learning impairs subsequent memory recombination with past episodes

“…We tested the hypothesis that novel threat learning may strengthen associated elements of a pre-existing, reactivated memory, in line with earlier observations (Zhu et al, 2020). Moreover, we tested whether this effect was moderated by the anticipatory arousal responses that the neutral B items came to evoke following B ® C learning, operationalized as the pupil response to B during the third round of learning.…”

“…Applying the same modelling strategy used to analyze the effect of threat learning and arousal on relational memory, we tested whether the original memories (A ® B) were strengthened by the novel threat associations (B ® C), which would be in line with earlier studies finding evidence of emotional tagging (Dunsmoor et al, 2015;Zhu et al, 2020).…”

“…For example, recognition memory has been shown to be enhanced for items that, through new emotional learning, have acquired some predictive value as a semantic category (Dunsmoor et al, 2015). As for associative memories, Zhu et al (2020) recently revealed that when one element of an existing memory is paired to an emotional stimulus, associations between elements of that original memory are strengthened. Similar evidence comes from a study using monetary reward as a reinforcing stimulus, showing that a hippocampus-dependent mechanism allows positive value to spread to older, indirectly related memories that are reactivated during reinforcement learning, biasing subsequent decision making (Wimmer & Shohamy, 2012).…”